Re: [femm] Iron losses in PM synchronous machine

[David Meeker <dmeeker@xxxxxxxxxxxxxxxxx>]

vecerkaj@xxxxxxx wrote:

> Hi everybody I use FEMM for field analyses of PM synchronous motors
> (the stator consists from standard three phases winding embedded in
> slots, a set of permanent magnets NdFeB are on the rotor). Does
> anybody have a suggestion, how to calculate the stator iron losses,
> when the rotor rotates by desired rpm? Is FEMM even able to solve such
> a type of problem? Till now, I use FEMM for the calculation of torque
> characteristics and total flux and the results are quite accurate in
> comparison with data abtained by measurements on a real motor. But the
> iron losses, it's different story and I have no idea how to solve it.
> I'll be glad for any response, suggestion, advice. Thanks in advance.
> George Vecerka Department of Power Electrical and Electronic
> Engineering Faculty of Electrical Enginnering and Computer Science
> Brno University of Technology Czech Republic

Sorry that I haven't been able to address this one for a while--I've
been sort of busy. This sort of problem can be addressed via femm, but
it's not exactly straightforward. You end up splitting up the problem
into multiple steady-frequency harmonic problems, summing the losses
from all subproblems to get the total losses.

First, I assume that you have some sort of surface-mount PM machine.
Second, I would assume that it is OK to assume that your magnets have a
relative permeability of exactly 1. In this case, it is possible to
think of the permanet magnets as simply the equivalent of a set of line
currents (like in section A1 of the manual). You then analytically
formulate a fourier series that replaces these currents on the edges of
the permanent magnets with a set of sinusoidally distributed current
densities at different spatial frequencies. When you spin the rotor at
a constant speed, you find that the frequency experienced by the stator
due to the permanent magnet currents is different for each harmonic in
the fourier series. Therefore, you end up modeling the PM magnets in
femm as if they were an air-core winding on the motor, picking the
amplitude and phase of the currents in the "winding" so that the
currents appear to rotate around the rotor. For the fundamental, you
have contributions from both the stator currents and the fundamental of
the magnet currents. For the higher harmonics, you have just the
contribution from the magnet currents. Specify the "right" lamination
thickness and materials properties for the stator laminations, run cases
for the fundamental and "enough" higher harmonics. For each run,
evaluate the "total losses" integral over the core to get the hysteresis
and eddy current losses in the core. To get the total losses, just sum
up the results from all harmonics. You know that you have "enough"
harmonics when the loss result converges.

This is probably an impossibly terse description of this method--I
haven't had time to write up anything more elaborate, but I would be
glad to answer questions on it if you are still interested.

Dave Meeker



begin:vcard 
n:Meeker;David
tel;fax:781-890-3489
tel;work:781-684-4070
x-mozilla-html:TRUE
url:http://members.aol.com/dcm3c
org:Foster-Miller, Inc.;Electrical and Electronic Systems Group
version:2.1
email;internet:dmeeker@xxxxxxxxxxxxxxxxx
title:Senior Engineer
adr;quoted-printable:;;350 Second Avenue=0D=0A;Waltham;MA;02451-1196;USA
fn:David Meeker, Ph.D.
end:vcard